NexusAI: Multimodal Collaborative Research Swarm

Already contacted

This project builds a voice-driven intelligent collaborative research swarm, solving pain points of tedious interaction and scattered information in complex research. Users send one voice command, and the system auto-completes intent recognition, web-wide information gathering, deep analysis, and report generation. Core innovation: anthropomorphized Swarm collaboration. We gave Agents distinct personalities (like 'leader-type' Router, 'show-off' intel officer, 'deep' analyst). They flow data through event buses in backend while communicating and praising like a real team in frontend, achieving experience leap from 'tool usage' to 'human-machine symbiosis.'

Planning 3 core Agents using Hub-and-Spoke (star) collaboration mode: Router (Commander) integrates ASR tools, handles intent recognition, task distribution and anti-echo control as team Leader. Web-Searcher (Intel Officer) integrates search engines for breadth search and cleaning. Analyst handles long-text reading and deep reasoning, outputting structured reports. Collaboration through OpenAgents event system. Agents don't dialogue directly, exchanging data via standardized task.delegate and task.complete events. Uses openagents.mods.workspace.project module for 'speaking publicly' in group chat, achieving dual-track parallel of business logic (Event) and interactive feedback (Message).

① Router Agent (Multimodal Hub): Audio stream processing, auto-recognizing .wav files and transcribing; intelligent routing filtering system echoes, precisely recognizing intent and dispatching downstream Agents. ② Web-Searcher Agent (Active Intel): Web search capability (Google/HackerNews), preliminary summarization; 'announce before deliver' logic proactively reporting progress before submitting data. ③ Analyst Agent (Deep Synthesis): Cross-document extraction and viewpoint aggregation, generating Markdown reports with 'core insights', 'key intelligence' and 'expert perspectives'. ④ Fully automated workflow: voice input → transcription → intent distribution → parallel search → deep analysis → archival. ⑤ Anthropomorphic Team Atmosphere (Social Mod): Workplace interactions between Agents via Prompt Engineering and triggers.